Complete Factorial Design for Optimization of Operating Conditions for a Nanofiltration 90 Polymeric Membrane Treating High Concentration Sulfated Waters and Modeling Using Machine Learning
Bibliographic record
Abstract
In the prairie provinces of Western Canada, including Saskatchewan, many farms lack access to potable, healthy water and rely on dugouts for their water supply. Dugouts are artificial ponds or reservoirs that collect and store water, often from rain or snowmelt, for agricultural and livestock use. Dugouts contain, in some cases, high sulfate con-taminants that impact livestock watering. To clean these kinds of waters, a full factorial design study with eleven experiments was carried out to evaluate and optimize key nanofiltration membrane operating conditions, such as Trans-Membrane Pressure (TMP), Crossflow Velocity (CVF), and magnesium sulfate (MgSO4) concentration, fo-cusing on their impact on rejection rates and permeate flux. With optimal conditions of a TMP of 9 bar and a CFV of 0.65 m/s, the nanofiltration (NF90) membrane achieved a sulfate rejection of 90% and a permeate flux of 127 LMH, with CFV identified as the most significant factor influencing the operation of the membrane at all concentrations. Analysis of Variance (ANOVA) confirmed the statistical significance of the polynomial regression models, with a 95% confidence interval (CI). The membrane's rejection data and flux regression models yield a strong fit to the data, with a correlation coefficient exceeding 99%. Using the experimental dataset, two machine learning algorithms— Decision Tree (DT) and Random Forest (RF) — were employed to predict the permeate flux. The RF model demonstrated excellent predictive performance across all data sets, achieving a root mean square error (RMSE) of 3.98 and a coefficient of determination (R²) of 0.99. These findings highlight the potential of machine learning for predicting effec-tive sulphated water treatment.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".